S TAT E O F T H E U N I O N
income inequality
The Stanford Center on Poverty and Inequality
BY JONATHAN FISHER AND TIMOTHY M. SMEEDING
KEY FINDINGS
• When disposable-income
inequality is measured across
20–35 years of survey data,
the consistent result is that
the U.S. has the highest
level of disposable-income
inequality among rich
countries.
• The U.S. has the largest
“social distance” gap
between households at
the 90th percentile and
households at the 10th
percentile in the income
distribution. The U.S. has
the highest 90th percentile
point, meaning our rich are
indeed better off than those
in other countries, as almost
everyone expects. But our
poorest, at the 10th percentile,
are also lower in real terms
than are the poor in all other
comparison countries save
Italy.
• Some countries have
experienced periods of
falling, as well as rising,
inequality over the last three
decades. The simple, but
important, conclusion to
draw is that rising income
inequality is not inevitable.
Policy and markets can both
make a difference.
T
he explosion of research on income
inequality in the United States has
uncovered key facts about the sources and
patterns behind the takeoff. We know, for
example, that the long-term trend in income
inequality has been driven by two main factors: a surge at the top end in income and
wealth; and, at the bottom end, a combination of reduced wealth and slower income
growth during good times and a fall in
income during bad times. We likewise know
a lot about the role of education, technology,
deunionization, and globalization in bringing
about the takeoff in income inequality in the
U.S.
Although there has also been much crossnational comparative research on income
inequality, this line of research is somewhat less well-known and will therefore be
the focus of our article. The comparative
approach works well to expose the distinctiveness of the U.S. We live, of course, in a
famously exceptional country, but nowhere
is the U.S. more exceptional than in its level
of economic inequality.1
By examining cross-nationally comparable
measures of income inequality, we can move
beyond the often parochial debates about
U.S. inequality and come to appreciate how
our distinctive institutions create distinctive outcomes. Also, by comparing recent
trends in inequality across several nations,
we can better understand what U.S. policy
has and has not achieved and, more importantly, how it might be made more effective.
We can also better understand the effects
of extreme inequality on mobility, economic
growth, and other outcomes we value.
Measurement
We measure income inequality in terms
of disposable cash income (DPI), which is
adjusted for household size and (a) includes
all types of money income, (b) subtracts
out direct income and payroll taxes, and
(c) reflects all cash and near cash transfers,
such as food stamps, cash housing allowances, and refundable tax credits (e.g.,
the Earned Income Tax Credit [EITC]).2 We
measure inequality in DPI with the Gini coefficient and the ratio of the 90th to the 10th
percentiles in the distribution. The Gini coefficient measures inequality on a scale from
0 to 1, with higher numbers representing
greater inequality. We also measure redistribution by comparing the Gini for pre-tax
and transfer “market income” (MI) to that
of DPI. By calculating the Gini for market
income, we are able to “take out” the direct
role of the government, via taxes and transfers, in influencing inequality.
Our analyses are based on 10 rich nations
with well-established welfare states: the
U.S.; the Anglo-Saxon nations of Australia,
Canada, the United Kingdom, and Ireland;
France, Italy, and Germany; and two Scandinavian countries, Norway and Sweden. We
employ data from the LIS data set (formerly
the Luxembourg Income Study) and the
OECD Income Distribution Database (IDD).3
In an online appendix table, we expand
the analyses of MI and DPI inequality to 33
nations, including several middle-income
countries (Online Appendix Figure A1).
PATHWAYS • The Poverty and Inequality Report 2016
income inequality
Cross-National Differences in Absolute and Relative
Inequality
We begin by comparing the level of inequality across our 10
countries for the latest year that is available in the LIS. We do
so for both market income and disposable personal income.
The countries in Figure 1 are ranked from low to high by the
Gini coefficient for disposable income. In all countries, the
Gini coefficient for disposable income exhibits less inequality than the Gini coefficient for market income, as taxes and
transfers redistribute income to lower-income households.
The difference between the Gini coefficients for market
income and disposable income, is in this sense, a measure of
the level of redistribution in each country. We find that the U.S.
and Canada have the least redistribution, but the disposableincome Gini for Canada is substantially below that for the U.S.
because Canadian market-income inequality is much lower.
As compared to other countries in Figure 1, we see that the
U.S. has very high market-income inequality, although Ireland
has yet higher market-income inequality and Italy, the United
Kingdom, France, and Germany have roughly the same level
of market-income inequality as the U.S. The U.S. nonetheless ends up with the highest disposable-income inequality
because it engages in relatively little redistribution. Although
33
Ireland has considerably more market-income inequality
than the U.S., it engages in substantially more redistribution
and thus ends up with substantially less disposable-income
inequality.
The U.S. has had the highest level of disposable-income
inequality among rich countries for some time. When disposable-income inequality is measured across 35 years of LIS
data or 20 years of IDD data, the consistent result is that the
U.S. has the highest level of disposable-income inequality
among rich countries, even when the comparison is extended
to include a more expansive set of countries than those in
Figure 1 (see Online Appendix Figure A1 for the expanded
comparison). As shown in Figure A1, only the middle-income
countries of Russia, Turkey, Mexico, and Chile have higher
disposable-income inequality than the U.S. The simple
conclusion: The U.S. is the world champion of disposableincome inequality among rich nations.
St
at
es
om
te
d
ni
U
U
ni
te
d
Ki
n
gd
y
Ita
l
lia
tra
a
Au
s
ad
an
e
C
nc
Fr
a
nd
la
Ire
y
an
m
en
G
er
ed
Sw
N
or
w
ay
0
.1
.2
.3
.4
Gini Coefficient
.5
.6
The foregoing measures of income inequality are not, of
course, affected by cross-national differences in mean
income. Critics of this relative approach to inequality often
argue that absolute living standards should also be taken
into account. Because the U.S. is richer than almost all other
OECD countries, those at a
given percentile in the income
FIGURE 1. Market and Disposable Income Inequality by Country Using the Gini Coefficient
distribution—say, the 10th percentile—may well be better off
in absolute terms than those
at the same percentile in other
rich countries.4
Disposable Income
Market Income
Note: 2012 or latest year available. Disposable income is market income plus transfers and minus taxes.
Source: LIS (formerly the Luxembourg Income Study) (LIS; www.lisdatacenter.org) and The Organisation for Economic Co-operation and
Development’s (OECD) Income Distribution Database (http://www.oecd.org/social/income-distribution-database.htm).
PATHWAYS • The Poverty and Inequality Report 2016
This proposition can be
assessed with LIS data by using
purchasing power parities to
convert all country incomes
into equivalent U.S. dollars.5
Using purchasing power parities allows us to compare real
levels of well-being at various
points in the income distribution across nations. For this
purpose, perhaps the most
basic measure of real levels of inequality is the decile
ratio, which shows the “social
distance” gap between the
household at the 90th percentile
and the household at the 10th
percentile in the income distribution.
34
income inequality
When this approach is taken, we find that the U.S. has the
highest decile ratio (see Figure 2). It also has the highest 90th
percentile point, meaning our rich are indeed better off than
those in other countries, as almost everyone expects. But our
poorest, at the 10th percentile, are also lower in real terms
than are the poor in all other countries save Italy.6 The high
decile ratio in the U.S. is generated, then, not just because
the rich are especially well off, but also because the poor are
especially poor. The combined effect is to generate a U.S.
decile ratio of more than 6, meaning the incomes of those
at the 90th percentile are, on average, six times higher than
are those at the 10th percentile.7 The next highest ratios are
below 5 in Australia and Canada, while the French, Swedes,
and Norwegians have decile ratios below 4. Indeed, our decile
ratio is roughly twice that of Norway, whose level of GDP per
person is just about the same as ours. The poor in Norway
(i.e., 10th percentile) enjoy more than twice the real incomes
of the poor at our 10th decile.
The U.S. is, therefore, the most unequal rich country on
earth, a conclusion that holds equally for absolute or relative
measurement. Were we always the most unequal? The next
section addresses this question.
90th percentile ($10,000s)
10th percentile ($10,000s)
2
4
6
90-10 Ratio
at
es
lia
St
d
ni
te
U
Au
st
ra
a
an
ad
ng
d
Ki
te
d
ni
U
C
om
ly
Ita
nd
la
Ire
y
m
an
er
G
Fr
an
ce
ed
en
Sw
N
or
w
ay
0
Ratios or Dollars (10,000s)
One lesson from Figure 3, also shown in Table 1, is that some
countries—in particular, Sweden, the United Kingdom, Ire-
Income Inequality by Country Using the 90th and 10th Percentiles in U.S. Dollars
8
FIGURE 2.
Trends in Income Inequality
Table 1 and Figure 3 present changes in disposable income
inequality from the 1980s through 2010, the earliest and latest
comparable information we have on these countries. We find
that the U.S., while starting from a very high level of inequality,
has also experienced one of the largest increases in inequality since 1979 (18%). The United Kingdom has seen a similar
rise (21%), and Australia (17%) and Germany (17%) are close
behind the U.S. (Table 1). But Sweden, for which data are first
available in 1981, has had the fastest increase in both absolute and relative inequality, though from a very low base. We
also see that France, Germany, and Sweden ended up with
roughly the same level of inequality in 2010, despite their very
different pathways since 1980 in achieving that level (Figure
3). Indeed, France has experienced an 11 percent decrease
in inequality (since 1978), while Germany and Sweden have
experienced an increase during that period. The U.S. data
suggest a return to rising income inequality after a recessioninduced pause, whereas post-2010 data for other nations are
not yet available for comparison.8
Note: Data are from 2010 or latest year available. Income definition is disposable household income from LIS. Income is in equivalent dollars using the square root of family size as the equivalence
scale. Data for all countries except Sweden come from LIS. The data for Sweden come from OECD, but the OECD does not report the levels for the 90th and 10th percentiles, just the ratio. Figures are
adjusted for purchasing power parity (PPP) differences using OECD PPP numbers.
Source: LIS; OECD.
PATHWAYS • The Poverty and Inequality Report 2016
income inequality
land, and France—have experienced periods of falling, as
well as rising, inequality over the last three decades. The simple, but important, conclusion to draw is that rising income
inequality is not inevitable. Policy and markets can both make
a difference.
But the dominant result is, of course, one of rising inequality. The descriptive source of this trend is twofold: at the top
end, there has been a surge in high incomes; and at the bottom end, there has been much slower income growth during
good times and, in some cases, a fall in income in bad times.
Further, changes in labor and capital markets since 2000 have
combined to narrow and shrink the middle class in the U.S.
and in other nations.9 In OECD countries, taxes and benefits
have historically been effective in reducing inequality, especially in the decade prior to the Great Recession. In the midst
of the Great Recession, benefits for the unemployed and other
redistribution measures managed to at least partially stem the
rise in inequality generated by the market. But now, as we
finally emerge from the Great Recession, the fear is that the
effect of taxes and benefits has become weaker, accelerating
the overall upward trend in disposable-income inequality. This
pattern is visible in some, but not all, rich countries.10
Trends in Gini Inequality Since 1978
First Year
Last Year
Absolute
Change
Percent
Change
United States
1979
2010
0.058
18.0%
Australia
1981
2010
0.050
16.7%
Canada
1981
2010
0.031
10.3%
France
1978
2010
-0.038
-11.4%
Germany
1981
2010
0.044
16.9%
Ireland
1987
2010
-0.026
-7.8%
Italy
1989
2010
0.007
2.3%
Norway
1979
2010
0.022
8.7%
Sweden
1981
2005
0.089
43.4%
United Kingdom
1979
2010
0.059
20.8%
Note: Income definition is disposable household income. Income is in equivalent dollars using
the square root of family size as the equivalence scale.
Source: LIS; OECD.
Trends in the Gini Coefficient Using Disposable Income
.4
FIGURE 3.
TABLE 1.
.35
United States (18% Increase)
United Kingdom (21% Increase)
Gini coefficient
.3
.25
35
France (11% Decrease)
Germany (17% Increase)
.2
Sweden (43% Increase)
1980
1990
2000
Note: Income definition is disposable household income. Income is in equivalent dollars using the square root of family size as the equivalence scale.
Source: LIS; OECD.
PATHWAYS • The Poverty and Inequality Report 2016
2010
36
income inequality
for large intergenerational transfers of wealth to assure the
social position and status of one’s children.14
Income and Wealth
We have concluded to this point that the U.S. is the “most
unequal rich country on earth.” Does this conclusion hold
when the focus shifts to wealth inequality? Although Gabriel
Zucman covers wealth inequality in detail in the next chapter
of this report, a brief head-to-head comparison of the wealth
and income results will be revealing here.11 The OECD has
garnered comparable wealth inequality data that allow us to
examine the income and wealth shares of the top 10 percent
in each of our major nations except Ireland (see Figure 4).12
Why Do We See These Patterns?
The data observed here suggest that inequality is high and rising in most rich nations—especially, but not only, in the U.S.
The drivers of these changes are rooted in several decades of
globalization of trade, technological change, and the growth
of both high- and low-end service jobs. Because labor markets have always been the heart of market income for middle
class households, the foregoing changes—all of which have
profound implications for the labor market—have had a major
impact on earnings and incomes in the U.S. and elsewhere.15
We clearly see that the U.S. is far and away the country with
the most unequal wealth distribution. The top 10 percent in
the U.S. have about 80 percent of all net worth, compared to
60 percent in Germany and Sweden, and 50 percent or less
in each of the other nations. It follows that the U.S. not only
has the weakest safety net but that the poor also have little
in the way of personal savings to cushion drops in income or
to meet unexpected expenses. The top decile, by contrast,
can easily self-insure against economic risks, afford its own
health care and education, and opt out of public sector provision of social goods.13 This amassing of wealth, which has
grown more unequal in recent years in the U.S., also allows
Income and Wealth Shares
90
FIGURE 4.
These changes have implications for those at both the top
and bottom of the labor market. At the top, the rise of highdemand business services and of winner-take-all markets
dominated by the well-educated have led to substantial
increases in earnings for the “winners,” thus leading to a
pulling-away at the top. Meanwhile, at the other end of the
labor market, wages of workers with low skills have not kept
up, and manual jobs in goods production and distribution
(manufacturing, assembly, and shipping) constitute a fall-
Top 10% Share - Wealth
Note: 2012 or latest year available.
Source: OECD. 2015. In It Together: Why Less Inequality Benefits Us All. Paris: OECD Press; Luxembourg Wealth Study (LWS).
PATHWAYS • The Poverty and Inequality Report 2016
St
at
es
m
U
d
ni
te
U
ni
te
d
Ki
n
gd
o
Fr
an
ce
Ita
ly
lia
Au
st
ra
ad
a
an
C
an
y
G
er
m
en
ed
Sw
N
or
w
ay
0
10
20
30
Percent
40 50
60
70
80
Top 10% Share - Income
income inequality
ing proportion of total employment. In some countries (esp.
Germany), “permanent employment” labor market institutions have been able to forestall widespread unemployment
in these sectors, even since the Great Recession. But even in
these countries, there are rises in inequality due to changes
at the high end of the earnings scale.16 Because the U.S. is
built around a low-wage market, with relatively high numbers
of short-term and part-time jobs, the changes at the bottom
of the labor market are especially important.17
We have also seen changes in income tax systems that have
reduced marginal tax rates for high earners. According to
the OECD,18 taxes and benefits have tended to redistribute
less in the period from the mid-1990s up to the Great Recession. It follows that rising inequality is generated not just by
institutional changes in labor markets, but also by declining
redistribution.
Where to Go from Here?
As long as the U.S. relies almost exclusively on the job market to generate incomes for working-age families, economic
changes that reduce the earnings of less-skilled workers will
inevitably have a big negative effect on inequality among children and prime-age adults, the vast majority of whom have
little wealth. This means that if the U.S. wishes to reduce
inequality, it can either (a) alter its labor market institutions to
ensure that there are more workers and that workers are paid
better, or (b) alter its redistributive institutions to reduce its
reliance on the job market.
If the U.S. fails to alter its policies in either of these ways,
the implications of such inaction would seem to be clear. The
high direct and indirect costs of inequality are now becoming widely recognized in public debates, both nationally and
cross-nationally.19 Because of high inequality, U.S. economic
growth and human capital growth have been lower and far
below expectations. And social mobility in the U.S. is much
lower than in other rich nations (as Miles Corak discusses in
his chapter).20 Although many economists favor globalization
and free trade because—in the aggregate—the gains from
37
trade exceed the losses, the key problem under this formulation is developing institutions that allow the many winners
(who pay lower prices for higher quality goods) to compensate the losers (whose jobs are lost to imports). The obvious
point here is that—to date—the winners have not compensated the losers.
The further worry is that a high-inequality economy reduces
the amount of social mobility and opportunity. As many have
argued, inequality can affect growth and upward mobility by
reducing educational opportunities for children from poor and
lower-middle-class families, thus lowering their future earnings and incomes. It follows that inequality reduces social
mobility and overall growth because of slow skill development.21 In the U.S., rich parents provide a “private safety net”
for their children, thus circumventing the problems arising
from low levels of public support for education, health care,
or other institutions. This private safety net allows rich children to exploit opportunities and thus reduces mobility and
economic growth.
Up to now, the U.S. has shown its indifference to high and
rising levels of inequality, although the outcry for change has
recently grown louder and more insistent. This brief report
has made it clear that ever-rising inequality is not inevitable,
that declining mobility is not inevitable, and that rich countries have in fact made choices about their labor-market
policies and their levels of redistribution. Although the high
levels of inequality in the U.S. are the residue of past choices,
there have of course been historic moments in U.S. history in
which new pathways have been charted and new policies and
institutions have been introduced. We cannot rule out that a
moment of this sort is nearing in which the U.S. adopts more
progressive policies that reduce inequalities and promote the
general welfare. ■
Jonathan Fisher is Research Scholar at the Stanford Center on
Poverty and Inequality. Timothy M. Smeeding is Arts & Sciences
Distinguished Professor of Public Affairs and Economics at the
University of Wisconsin.
PATHWAYS • The Poverty and Inequality Report 2016
38
income inequality
NOTES
1. On American exceptionalism, see Kenworthy, Lane. 2015. “America is Exceptional… and
Ordinary.” The Good Society. Available at http://
lanekenworthy.net/america-is-exceptional/.
2. In-kind transfers in the form of health and
education benefits are not counted, nor are
indirect taxes. For an analysis of cross-national
income inequality that includes these items, see
Garfinkel, Irwin, Lee Rainwater, and Timothy
Smeeding. 2006. “A Reexamination of the Welfare State and Inequality in Rich Nations: How
In-Kind Transfers and Indirect Taxes Change the
Story.” Journal of Policy Analysis and Management 25(4), 855–919.
3. Cross-National Data Center in Luxembourg,
available at http://www.lisdatacenter.org/;
OECD Income Distribution Database (IDD):
Gini, Poverty, Income, Methods and Concepts,
available at http://www.oecd.org/social/incomedistribution-database.htm.
4. For instance, see Wilkinson, Will. 2009.
Thinking Clearly about Economic Inequality.
Policy Analysis, No 640. Washington, D.C.:
Cato Institute. Available at http://object.cato.
org/sites/cato.org/files/pubs/pdf/pa640.pdf.
5. Purchasing power parities (PPP) allow comparison of incomes across countries. Thus we
are able to convert all income to U.S. dollars
in order to compare incomes across countries.
Our PPP data come from the OECD.
6. All data in Figure 2 come from LIS, except
for Sweden, which comes from the IDD. The
IDD only publishes the 90-10 ratio and not the
actual values at the 90th percentile and 10th
percentile.
7. See also Keeley, Brian. 2015. Income
Inequality: The Gap between Rich and
Poor. OECD Insights. Paris: OECD Publishing. Available at DOI: http://dx.doi.
org/10.1787/9789264246010-en. He finds
even bigger differences by examining average
incomes in the top and bottom deciles.
8. LIS key figures, available at http://www.
lisdatacenter.org/lis-ikf-webapp/app/search-ikffigures.
9. Organisation for Economic Co-operation
and Development (OECD). 2015. In It Together:
Why Less Inequality Benefits All. Paris: OECD
Publishing. Available at DOI: http://dx.doi.
org/10.1787/9789264235120-en; Gornick,
Janet, and Branko Milanovic. 2015. Income
Inequality in the United States in Cross-National
Perspective: Redistribution Revisited. LIS Center Research Brief (1/2015). Available at https://
www.gc.cuny.edu/CUNY_GC/media/CUNYGraduate-Center/PDF/Centers/LIS/LIS-CenterResearch-Brief-1-2015.pdf; and Pew Research
Center. 2015. “The Lost Decade of the Middle
Class: Fewer, Poorer, Gloomier.” Washington,
D.C.: Pew Research Center. Available at http://
www.pewsocialtrends.org/2012/08/22/the-lostdecade-of-the-middle-class/.
10. OECD, 2015.
11. See Fisher, Jonathan, David Johnson, Jonathan Latner, Timothy Smeeding, and Jeffrey
Thompson. 2016. “Inequality and Mobility Using
Income, Consumption, and Wealth for the Same
Individuals.” Russell Sage Foundation Journal
of the Social Sciences, forthcoming.
12. The reader should be aware that the top
income and wealth shares are each calculated
from different databases, so that the top 10 percent are different households in each database.
Fisher et al. (2016), in contrast, examine wealth
income and consumption in the U.S. for the
same persons.
13. Osberg, Lars, Timothy Smeeding, and
Jonathan Schwabish. 2004. “Income Distribution and Public Social Expenditure: Theories,
Effects, and Evidence.” In Social Inequality
(K. Neckerman, Ed.). New York: Russell Sage
Foundation.
14. Yellen, Janet L. 2014. “Perspectives on
Inequality and Opportunity from the Survey
of Consumer Finances.” Presentation to the
Conference on Economic Opportunity and
Inequality, Federal Reserve Bank of Boston.
October 17, 2014. Boston. Available at http://
federalreserve.gov/newsevents/speech/yellen20141017a.htm (last accessed December 5,
2015); Fisher et al. (2016).
15. Autor, David. 2014. “Skills, Education, and
the Rise of Earnings Inequality Among the ‘Other 99 Percent’.” Science 344(1866), 843–851.
16. Haveman, Robert, Carolyn Heinrich, and
Timothy Smeeding. 2011. “Policy Responses
to the Recent Poor Performance of the United
States Labor Market.” Journal of Policy Analysis
and Management, 31(1), 177–186.
17. National Employment Law Project. 2012.
The Low-Wage Recovery and Growing Inequality. Data Brief. New York: NELP. Available at
http://www.nelp.org/content/uploads/2015/03/
LowWageRecovery2012.pdf.
18. OECD, 2015.
19. OECD, 2015; Organisation for Economic
Co-operation and Development (OECD). 2014.
“Focus on Inequality and Growth – December 2014.” Paris: OECD. Available at http://
www.oecd.org/social/Focus-Inequality-andGrowth-2014.pdf; Kenworthy, Lane, and Timothy Smeeding. 2014. “The United States: High
and Rapidly Rising Inequality.” Inequality and
Its Impacts, Volume 2 (Brian Nolan and Wiemer
Salverda, et al., Eds.). New York: Oxford University Press.
20. See also Corak, Miles. 2013. “Income
Inequality, Equality of Opportunity, and Intergenerational Mobility.” Journal of Economic
Perspectives 27(3), 79–102.
21. Figure 1.6 in OECD, 2015.
PATHWAYS • The Poverty and Inequality Report 2016